The U.S. government’s export control order has forced Anthropic to disable its advanced AI models for foreign nationals, significantly increasing interest in Canadian AI company Cohere. On June 12, the Trump administration’s Commerce Department mandated that Anthropic suspend access to its Fable 5 and Mythos 5 models for all foreign nationals, including the company’s own employees outside the U.S. This unprecedented action, citing national security concerns, marked the first time the government applied export controls directly to AI models rather than semiconductor hardware. Following the directive, Anthropic sent staff to Washington to negotiate a resolution.
Cohere, based in Toronto, has positioned itself as a sovereign alternative to U.S. AI providers. Joelle Pineau, Cohere’s chief AI officer, reported a significant increase in inquiries from business customers and governments seeking to diversify their technology options. Pineau stated, “Governments are concerned about their own ability to access the technology,” emphasizing that the ban created volatility unacceptable to enterprise and government clients. Cohere’s offering of on-premises deployment allows customers direct control over their AI infrastructure, enhancing its appeal in light of recent developments.
The situation has also led to strengthened ties between Cohere and Canadian defense, exemplified by a new partnership with Calian Group to integrate sovereign AI into military settings. Additionally, the Canadian government is pursuing a $2 billion Sovereign AI Compute Strategy aimed at building domestic AI capabilities.
The Anthropic ban has triggered broader discussions regarding AI sovereignty among allied nations. The Globe and Mail highlighted that the directive exposed vulnerabilities in Canada’s reliance on American AI firms, while experts noted that under the export control regime, Canadians are considered foreign nationals. In Europe, the suspension of Anthropic’s models has similarly prompted calls for greater technological independence from U.S. providers. Fortune described the situation as evidence that the U.S. now effectively manages a licensing regime for advanced AI models that is both “ad hoc and opaque.”
